In an earlier report, we showed that cells lacking the ARH1 gene showed enhanced proliferation compared to ARH1-/- cells transformed with a wild-type ARH1 gene;cells transformed with a mutant ARH1 gene (D60A/D61A), which encodes a protein that exhibits less than 0.1 percent of wild-type activity, also showed enhanced proliferation similar to ARH1-/- cells. ARH1-/- cells also produced colonies in soft agar, as did ARH1-/- cells transformed with the ARH1 gene containing the double mutation. In contrast, ARH1-/- cells transformed with the wild-type gene did not show colonies in soft agar. In further support for the tumorigenic potential of ARH1-/- cells, ARH1-/- cells and cells transformed with the mutated ARH1 gene also formed tumors in nude mice. In contrast, ARH1-/- cells transformed with the wild-type gene did not give rise to tumors when injected into nude mice. Of importance, ARH1+/- cells from heterozygous animals generated tumors in nude mice. In some heterozygous ARH1+/- mice, we also observed tumor development. We examined the functional allele from the tumors in nude mice injected with ARH1+/- cells or in tumors in heterozygous animals. In some instances, the functional allele appeared to be absent, consistent with loss of heterozygosity (LOH). These data are compatible with the autosomal dominant model of tumorigenesis proposed by Alfred Knudson, in which LOH occurs as the second hit leading to tumor formation. In other instances, however, there appeared to be an intact ARH1 allele in the heterozygote. The ARH1 gene from DNA isolated from those tumors was sequenced and contained nucleotide differences from the ARH1 gene in non-tumor DNA from nude mouse or in DNA from non-tumor tissue isolated from the heterozygote mouse. The mutant genes exhibited changes in exons 2 and 3, which are part of the coding region and may be responsible for the structure of the catalytic site and thus be expected to alter enzymatic activity. We next looked at the sequences of the mutant ARH1 gene from heterozygote cells injected into nude mice and from tumor obtained from heterozygote animals. To understand better the effects of the mutation, recombinant proteins were synthesized from the mutant genes and their activities were measured in a standard assay where the hydrolysis of ADP-ribose-arginine was quantified. In most cases, the mutant protein exhibited enzymatic activity, which was a fraction of the wild-type protein, consistent with the fact that the mutation occurred in a coding region exon, probably associated with the active site. When ARH1 KO cells were transfected with the mutant genes and the resulting transformed cells were injected subcutaneously into nude mice, tumors were observed in a time-dependent manner, even though the gene could be used to synthesize a protein exhibiting enzymatic activity. Thus, the presence of enzymatic activity is not sufficient to prevent tumor development. Conceivably, the mutation might affect the ability of the protein to be properly localized or to use different ADP-ribosylated proteins as substrates. The tumor growth rate following subcutaneous injection of KO cells was faster than those of cells transformed with a mutant gene, encoding an active ARH1. Thus, in quantitative assays, the presence or absence of a mutation may affect tumor development. KO cells transformed with a wild-type ARH1 also developed tumors depending on the level of cell lysate ARH1 activity. Cells having relatively low levels of ARH1 activity developed tumors in nude mice. Of note, when normalized to cell ARH1 enzymatic activitiy, transformation with the wild-type gene was more effective than transformation with mutant genes in slowing the rate of tumor development. Thus, abnormalities in the ARH1 gene appeared to determine, in part, the tumorigenic potential of the ARH1 KO cells.